Aluminum-steel dissimilar metal needleless friction spot welding method based on multi-stage pressing

Abstract

This invention discloses a needle-free friction spot welding method for aluminum and steel dissimilar metals based on multi-stage pressing, belonging to the field of friction spot welding technology. The invention employs an overlapping structure with a steel plate placed on top of an aluminum plate, dividing the pressing process of the rotating tool into three continuous stages: initial frictional heat generation, interfacial diffusion reaction, and upsetting. Each stage is configured with appropriate displacement rate and pressure control logic. The initial stage is based on real-time torque dynamic adjustment parameters; the interfacial stage couples the real-time temperature field of the steel-aluminum interface with closed-loop pressure regulation; and the upsetting stage employs adaptive control with step-increased pressure. This invention can achieve precise matching between the pressing process and the interfacial metallurgical reaction, effectively controlling the growth of brittle phases at the interface, and improving the weld nugget formation quality and joint mechanical properties.

Description

Technical Field

[0001] This invention relates to the field of friction spot welding technology, and specifically to a needleless friction spot welding method for dissimilar metals such as aluminum and steel based on multi-stage pressure. Background Technology

[0002] Existing needle-free friction spot welding processes for dissimilar metals like aluminum and steel typically employ a single-stage, constant-parameter pressing scheme. The specific process involves: a steel plate overlapping an aluminum plate; a spindle driving a rotating tool downwards to contact the upper surface of the steel plate; and then the rotating tool pressing downwards at a constant displacement rate and constant axial pressure until a preset total pressing depth is reached. Throughout the pressing process, the rotational speed of the tool remains constant, and the axial pressure and displacement rate applied to the lap joint are fixed values, not dynamically adjusted based on temperature changes at the welding interface, the plastic flow state of the material, or phase transformation processes.

[0003] The aforementioned single-stage constant-parameter pressing scheme has a core flaw: the single-stage pressing process cannot adapt to the dynamic process of metallurgical reaction and plastic deformation at the steel-aluminum interface. Due to the significant differences in the physicochemical properties of aluminum and steel, the metallurgical reaction and plastic deformation at the weld interface exhibit highly nonlinear changes with increasing temperature. Using a fixed pressure and displacement rate cannot provide adequate frictional heat generation and oxide film removal in the initial stage, cannot adjust the heat input according to the interface temperature distribution to control the growth of intermetallic compounds in the intermediate reaction stage, and cannot provide a matching forging force to enhance the deformation resistance of the material in the later forming stage. This results in excessive generation of brittle phases at the interface and poor weld nugget forming stability. Summary of the Invention

[0004] The purpose of this invention is to provide a needleless friction spot welding method for dissimilar metals such as aluminum and steel based on multi-stage pressure, which can solve the problems in the background art mentioned above.

[0005] To achieve the above objectives, the technical solution adopted by the present invention is as follows: A needle-free friction spot welding method for aluminum and steel dissimilar metals based on multi-stage pressure reduction includes: A steel plate is placed on top of an aluminum plate to form a lap joint, and a rotating tool is started to frictionally heat the lap joint. The pressing process of the rotary tool is divided into an initial frictional heat generation stage, an interfacial diffusion reaction stage, and an upsetting stage. During the initial frictional heat generation stage, the rotating tool is pressed down according to the first displacement rate and the first dynamic pressure curve until the first preset displacement threshold is reached. During the interface diffusion reaction stage, the rotating tool is pressed down according to the second displacement rate and the second dynamic pressure curve until the second preset displacement threshold is reached. The output pressure value of the second dynamic pressure curve is coupled with the real-time distribution data of the temperature field at the steel-aluminum interface. During the upsetting stage, the rotary tool is pressed down to the final welding depth according to the third displacement rate and the step-increasing pressure curve, and a pressure holding operation is performed at the final welding depth.

[0006] Preferably, in the initial frictional heat generation stage, the generation process of the first dynamic pressure curve includes: acquiring the real-time compression depth value of the rotating tool, inputting the real-time compression depth value into a preset displacement-pressure mapping relationship model, and calculating the initial base pressure value; collecting the real-time torque value of the rotating tool during rotation, correcting the initial base pressure value based on the real-time torque value, and generating the first dynamic pressure curve; the first displacement rate is set as a nonlinear deceleration rate, and the decrease amplitude of the nonlinear deceleration rate is positively correlated with the fluctuation frequency of the real-time torque value.

[0007] Preferably, in the interface diffusion reaction stage, the method for acquiring the real-time distribution data of the steel-aluminum interface temperature field includes: extracting an infrared thermal radiation image sequence from the side of the lap joint at a preset sampling frequency; performing grayscale processing on the infrared thermal radiation image sequence to extract the grayscale feature matrix of the steel-aluminum interface; comparing the grayscale feature matrix with a pre-calibrated temperature grayscale reference table to generate real-time distribution data of the steel-aluminum interface temperature field; and adjusting the output pressure value of the second dynamic pressure curve in a piecewise linear inverse proportion to the change of the highest temperature value in the real-time distribution data.

[0008] Preferably, in the upsetting stage, the stepped pressure curve includes three consecutive pressure steps. The pressure value of the first pressure step is set to a first multiple of the instantaneous pressure value at the end of the interfacial diffusion reaction stage, the pressure value of the second pressure step is set to a second multiple of the instantaneous pressure value, and the pressure value of the third pressure step is set to a third multiple of the instantaneous pressure value. The first multiple, the second multiple, and the third multiple increase sequentially. The third displacement rate remains constant within the three consecutive pressure steps.

[0009] Preferably, before starting the rotating tool to perform friction heating on the lap joint, the method further includes: applying a pre-clamping force to the rotating tool so that the end face of the rotating tool contacts the surface of the steel plate; during the application of the pre-clamping force, controlling the rotating tool to rotate at an initial speed, and simultaneously recording the axial displacement offset of the spindle of the rotating tool; when the axial displacement offset of the spindle reaches a preset offset threshold, increasing the speed of the rotating tool to a set working speed, and simultaneously triggering the initial friction heating stage.

[0010] Preferably, the stage switching control method of the pressing process includes: during the initial frictional heating stage, continuously calculating the rate of change of pressing displacement per unit time; when the rate of change of pressing displacement is lower than a first rate of change threshold and continues for a first set time, generating a stage switching command; according to the stage switching command, switching the pressing state of the rotating tool from the initial frictional heating stage to the interface diffusion reaction stage; during the interface diffusion reaction stage, when the area ratio of the critical temperature region for plastic softening on the steel side in the real-time distribution data reaches a preset area ratio threshold, generating a second stage switching command and switching to the upsetting forming stage.

[0011] Preferably, the process of correcting the initial base pressure value based on the real-time torque value includes: extracting the high-frequency torque fluctuation component from the real-time torque value and calculating the root mean square value of the high-frequency torque fluctuation component; dividing the root mean square value by the average value of the real-time torque value to obtain the torque fluctuation coefficient; inputting the torque fluctuation coefficient into a preset pressure compensation function to calculate the pressure compensation amount; adding the pressure compensation amount to the initial base pressure value to generate the corrected pressure value, and the corrected pressure value constitutes the first dynamic pressure curve.

[0012] Preferably, before performing grayscale processing on the infrared thermal radiation image sequence, the method further includes: performing Gaussian filtering denoising processing on each frame of the infrared thermal radiation image sequence; for each frame of the denoised infrared thermal radiation image, using an edge detection operator to extract the contour line of the interface between the steel plate and the aluminum plate; defining a rectangular region of interest in each frame of the infrared thermal radiation image based on the interface contour line, and performing subsequent grayscale processing and grayscale feature matrix extraction operations only on the pixels within the rectangular region of interest.

[0013] Preferably, in each of the three consecutive pressure steps, the normal pressure value applied by the rotating tool to the lap joint and the real-time indentation depth value of the rotating tool are collected in real time; the ratio of the normal pressure value to the real-time indentation depth value is calculated to obtain the real-time deformation resistance value; when the increase of the real-time deformation resistance value compared with the deformation resistance value of the previous collection cycle exceeds the preset resistance increase threshold, a step jump command is triggered, and the next pressure step is entered according to the step jump command.

[0014] Preferably, the process of performing the pressure holding operation during the upsetting stage includes: during the maintenance of the third pressure step, using a second set time as a sliding window, continuously calculating the variance of all the real-time deformation resistance values ​​within the sliding window; comparing the variance with a preset variance threshold, and generating a pressure holding termination command when the variance is less than the preset variance threshold for a third consecutive set time; and controlling the rotating tool to unload the pressure and lift it upwards to disengage from the lap joint according to the pressure holding termination command.

[0015] Compared with the prior art, the beneficial effects of the present invention are as follows: 1. This invention divides the pressing process into an initial frictional heat generation stage, an interfacial diffusion reaction stage, and an upsetting stage. For each stage, corresponding first displacement rate and first dynamic pressure curve, second displacement rate and second dynamic pressure curve, and third displacement rate and step-increasing pressure curve are configured. By coupling the output pressure value of the second dynamic pressure curve with the real-time distribution data of the temperature field of the steel-aluminum interface, the control process of pressing displacement and pressure is matched with the metallurgical reaction process of the steel-aluminum interface, thus overcoming the defect that the single-stage pressing process cannot adapt to the dynamic process of the interface, resulting in poor weld nugget formation stability.

[0016] 2. By extracting the high-frequency torque fluctuation component from the real-time torque value, the root mean square value is calculated and converted into a pressure compensation amount to correct the initial base pressure value. Combined with a nonlinear deceleration rate that is positively correlated with the torque fluctuation frequency, the mechanical energy input in the initial frictional heat generation stage is adapted to the non-uniformity of the material state. Gaussian filtering denoising and edge detection operator extraction are combined with the interface contour line to delineate a rectangular region of interest. Only the gray-scale feature matrix of this region is extracted and compared with the temperature gray-scale reference table to obtain the temperature field data. The pressure is adjusted piecewise linearly and inversely based on the highest temperature value, realizing the closed-loop regulation of heat and pressure in the interface diffusion reaction stage. In the upsetting stage, the real-time deformation resistance value is obtained by calculating the ratio of the normal pressure value to the real-time indentation depth value. When the increase in the deformation resistance value exceeds the preset resistance increase threshold, a step jump is triggered. Combined with the calculation of the deformation resistance value variance by a sliding window, which is continuously less than the preset variance threshold, the pressure holding is terminated, making the pressure application in the upsetting process directly related to the actual deformation state of the material. Attached Figure Description

[0017] Figure 1 This is a flowchart of the overall process for needleless friction spot welding of dissimilar metals such as aluminum and steel based on multi-stage pressure. Figure 2 Flowchart for triggering control during the pre-compression stage and the initial welding stage; Figure 3 The flowchart shows the dynamic pressure and rate adaptive control process during the initial frictional crushing stage. Figure 4 Flowchart for temperature field detection and pressure closed-loop control during the interfacial diffusion reaction stage; Figure 5 Flowchart of adaptive jump control for stepped pressure during the upsetting stage; Figure 6 The flowchart shows the control process for determining the termination of the pressure holding operation and the exit control upon completion of welding. Detailed Implementation

[0018] The needleless friction spot welding method for aluminum and steel dissimilar metals based on multi-stage pressure described in this specific embodiment can be applied to the lap welding of aluminum and steel dissimilar metal plates. The thickness of the steel plate and the aluminum plate can be adapted according to the actual welding requirements. The rotating tool used in the needleless friction spot welding is a needleless stirring head with a flat end face. The end face diameter of the rotating tool can be set according to the weld nugget size requirements of the lap joint.

[0019] In one embodiment, a steel plate is placed on top of an aluminum plate to form a lap joint. The size of the lap area between the steel and aluminum plates must match the end face size of the rotating tool. The center of the lap area is coaxial with the spindle center of the rotating tool to ensure that the frictional heat from the rotating tool during welding can be evenly applied to the core area of ​​the lap joint. The rotating tool is activated to perform frictional heating on the lap joint. The rotating tool rotates around its own spindle, and the frictional heat generated by the contact between the end face and the surface of the steel plate during rotation serves as the core heat source for the welding process. The frictional heat is conducted through the steel plate to the steel-aluminum interface, gradually achieving plastic softening and metallurgical reaction of the material at the interface.

[0020] refer to Figure 1 The pressing process of the rotating tool is divided into an initial frictional heat generation stage, an interface diffusion reaction stage, and an upsetting stage. The three stages are executed sequentially along the pressing direction of the rotating tool. Each stage corresponds to an independent displacement rate control logic and pressure control logic. The control logic of each stage forms a corresponding adaptation relationship with the changes in the physical state of the material and the metallurgical reaction process during the aluminum-steel dissimilar metal welding process.

[0021] During the initial frictional heat generation stage, the rotating tool is pressed down according to a first displacement rate and a first dynamic pressure curve until a first preset displacement threshold is reached. The first preset displacement threshold is related to the thickness of the steel plate and its value does not exceed 20% of the steel plate thickness. This ensures that the pressing process during the initial frictional heat generation stage only acts on the upper region of the steel plate. The core objective is to uniformly remove the oxide film on the steel plate surface, providing a fresh metal contact interface for subsequent interfacial metallurgical reactions. The first displacement rate is the dynamic change rate adapted to the initial frictional process, and the first dynamic pressure curve is a pressure sequence adjusted in real time during the pressing process. The two work together to match the initial stage frictional heat input with the oxide film removal process.

[0022] During the interfacial diffusion reaction stage, the rotating tool is pressed down according to the second displacement rate and the second dynamic pressure curve until a second preset displacement threshold is reached. The output pressure value of the second dynamic pressure curve is coupled with the real-time distribution data of the steel-aluminum interface temperature field. The value of the second preset displacement threshold does not exceed 40% of the steel plate thickness to avoid direct contact between the end face of the rotating tool and the aluminum plate, which could cause interfacial defects. The value of the second displacement rate is lower than the value of the first displacement rate at the end of the initial frictional heat generation stage, ensuring the stability of the pressing process and providing sufficient reaction time for interfacial atomic diffusion. The output pressure value of the second dynamic pressure curve is synchronously adjusted with the real-time changes in the steel-aluminum interface temperature field. By adjusting the pressure, the intensity of the frictional heat input is changed, thereby achieving closed-loop control of the interface temperature and constraining the growth process of intermetallic compounds at the interface.

[0023] During the upsetting stage, the rotating tool is pressed down to the final welding depth according to the third displacement rate and the stepped increasing pressure curve, and a pressure holding operation is performed at the final welding depth. The final welding depth is the preset maximum pressing depth, which does not exceed 60% of the steel plate thickness, ensuring that the rotating tool never directly contacts the aluminum plate throughout the welding process. The value of the third displacement rate is lower than the value of the second displacement rate at the end of the interface diffusion reaction stage to avoid excessive material spatter during upsetting. The stepped increasing pressure curve is a phased and gradually increasing pressure control sequence to provide a matching upsetting force for the densification of the weld nugget. The pressure holding operation is used to maintain the pressure environment at the interface, eliminate forming defects inside the weld nugget, and complete the final forming of the entire welding process.

[0024] Table 1. Stage Division and Core Control Parameter Benchmark Table for Multi-Stage Pressing Process Phase Name Displacement rate control mode Pressure control mode Displacement threshold setting benchmark Core control objectives Initial frictional heat generation stage Dynamically adjustable rate control Dynamic pressure control based on displacement and torque Not exceeding 20% ​​of the steel plate thickness. The oxide film on the steel plate surface is uniformly broken down, forming a fresh metal contact interface. Interfacial diffusion reaction stage Smooth speed control Dynamic pressure control coupled with interface temperature field Not exceeding 40% of the steel plate thickness. Controlling the thermal input at the interface enables the controlled diffusion reaction of steel and aluminum atoms. Upsetting Forming Stage Constant rate control Stepped increasing pressure control Not exceeding 60% of the steel plate thickness. Densification of the weld nugget eliminates internal defects and stabilizes joint performance. The table above clarifies the benchmark settings of the core control dimensions for the three consecutive pressing stages, providing a basic reference for the execution of control logic in each stage. This ensures that the pressing process in each stage is precisely matched with the material state changes and metallurgical reaction process in the aluminum-steel dissimilar metal welding process, avoiding the problem of mismatch between control parameters and interface dynamic process in the single-stage constant parameter pressing process.

[0025] In this embodiment, by dividing the pressing process of the rotary tool into three consecutive execution stages, and configuring corresponding displacement rate control logic and pressure control logic for each stage, the mechanical energy input of the pressing process is matched with the plastic deformation of the steel-aluminum interface and the metallurgical reaction process. This solves the problem that the single-stage constant parameter pressing scheme cannot adapt to the dynamic changes of the steel-aluminum interface, and provides a complete execution framework for the process control of needleless friction spot welding of dissimilar metals such as aluminum and steel.

[0026] refer to Figure 2 In a preferred embodiment, before initiating friction heating of the lap joint using a rotating tool, a pre-clamping force is applied to the rotating tool, causing its end face to contact the surface of the steel plate. The pre-clamping force is set below the minimum pressure value during the initial frictional heat generation stage to prevent excessive plastic deformation of the steel plate during the pre-clamping stage, while ensuring stable physical contact between the end face of the rotating tool and the steel plate surface, eliminating assembly gaps between them. During the application of the pre-clamping force, the rotating tool is controlled to rotate at an initial speed, which is lower than the set operating speed for subsequent welding processes, to prevent excessive frictional heat during the pre-clamping stage from causing premature softening of the steel plate. Simultaneously, the axial displacement of the rotating tool's spindle is recorded. This axial displacement is the axial displacement value of the rotating tool's spindle under the pre-clamping force, and is acquired in real-time by a displacement sensor built into the spindle at a preset sampling frequency of not less than 1 kHz to ensure the accuracy of displacement detection. When the axial displacement of the spindle reaches the preset offset threshold, it indicates that the end face of the rotating tool and the surface of the steel plate have formed a stable contact state of complete fit. At this time, the rotation speed of the rotating tool is increased to the set working speed, and the initial friction heat generation stage is triggered simultaneously to ensure that the initial state of the welding process is highly consistent and to eliminate the difference in the initial state caused by the assembly gap between different welding batches.

[0027] refer to Figure 3In the initial frictional heat generation stage of this embodiment, the generation process of the first dynamic pressure curve includes: acquiring the real-time pressing depth value of the rotating tool. The real-time pressing depth value is the axial displacement value of the end face of the rotating tool relative to the initial upper surface of the steel plate. This value is collected in real time by a displacement sensor built into the spindle, and uses the same reference coordinate system as the displacement detection in the pre-pressing stage to ensure the continuity and accuracy of the pressing depth value. The real-time pressing depth value is input into a preset displacement-pressure mapping relationship model to calculate the initial base pressure value. The displacement-pressure mapping relationship model is the correspondence between the pressing depth and the base pressure determined in advance through calibration tests. Its core logic is to gradually increase the base pressure value as the pressing depth increases, adapting to the material plastic deformation resistance that increases with the pressing depth. Specifically, the initial base pressure value is calculated using the following formula:

[0028] in, This represents the initial base pressure value corresponding to the real-time compression depth h. This is the displacement-pressure mapping coefficient, which is a pre-calibrated constant value. This represents the real-time depth of the rotary tool. The initial reference pressure value is the pre-tightening force value at the end of the pre-tightening stage, ensuring the continuity of the pressure control process.

[0029] The real-time torque value of the rotating tool during rotation is collected by a torque sensor built into the spindle at a preset sampling frequency of no less than 1 kHz to ensure that details of torque fluctuations are fully captured. Changes in the real-time torque value directly reflect the friction state between the end face of the rotating tool and the steel plate, the plastic flow state of the material, and the progress of oxide film removal. The initial base pressure value is corrected based on the real-time torque value to generate the first dynamic pressure curve. Through real-time correction of the torque value, the pressure control can adapt to the non-uniformity during the oxide film removal process, ensuring complete removal of the oxide film.

[0030] Specifically, the process of correcting the initial base pressure value based on the real-time torque value includes: extracting the high-frequency torque fluctuation component from the real-time torque value. This high-frequency torque fluctuation component is obtained by high-pass filtering the real-time torque value. The cutoff frequency of the high-pass filter is set to be at least five times the rotation frequency of the rotating tool. Low-frequency torque components synchronized with the rotation are filtered out, while the high-frequency fluctuation component generated during the removal of the oxide film on the steel plate surface is retained. The intensity of this high-frequency fluctuation component directly corresponds to the severity of the oxide film removal. The root mean square value of the high-frequency torque fluctuation component is calculated using the following formula:

[0031] in, This represents the root mean square value of the high-frequency torque fluctuation component. This represents the number of torque sampling points within a single calculation cycle. The value of the high-frequency torque fluctuation component at the i-th sampling point is set to a single calculation cycle between 10ms and 50ms to ensure the real-time performance of the calculation results.

[0032] The torque fluctuation coefficient is obtained by dividing the root mean square value by the average value of the real-time torque values, specifically through the following formula:

[0033] in, This is the torque ripple coefficient. The torque fluctuation coefficient is the average value of the real-time torque value within a single calculation cycle. The magnitude of the torque fluctuation coefficient directly reflects the degree of fluctuation of the current friction state. The larger the coefficient value, the more intense the oxide film removal process and the more uneven the friction state of the material.

[0034] The pressure compensation amount is calculated by inputting the torque fluctuation coefficient into a preset pressure compensation function, specifically through the following formula:

[0035] in, This is the pressure compensation amount. This is the torque-pressure compensation coefficient, which is a pre-calibrated constant value. Its value is positive, ensuring that the greater the torque fluctuation coefficient, the higher the pressure compensation. By increasing the pressure value, the frictional heat input is increased, accelerating the oxide film removal process.

[0036] The corrected pressure value is generated by adding the pressure compensation amount to the initial base pressure value, specifically through the following formula:

[0037] in, The corrected pressure value is formed by the sequence of changes in the corrected pressure value over time, which constitutes the first dynamic pressure curve. This dynamic pressure curve can be used to adapt to the changes in friction state during the oxide film removal process in real time, ensuring the uniform and complete removal of the oxide film.

[0038] In this embodiment, the first displacement rate is set as a nonlinear deceleration rate, and the decrease amplitude of the nonlinear deceleration rate is positively correlated with the fluctuation frequency of the real-time torque value. The fluctuation frequency of the real-time torque value is obtained by performing a fast Fourier transform on the real-time torque value, and the peak frequency in the high-frequency range is taken as the fluctuation frequency value. The level of the fluctuation frequency directly reflects the frequency of oxide film breakdown and the degree of inhomogeneity of the material state. The nonlinearly decreasing first displacement rate is specifically achieved by the following formula:

[0039] in, Let be the first displacement velocity at time t. This represents the initial displacement rate at the start of the initial frictional heat generation stage. This is the rate decay coefficient, which is a pre-calibrated constant value. The formula represents the fluctuation frequency of the real-time torque value. This formula achieves a non-linear decrease in the first displacement rate with running time. The higher the torque fluctuation frequency, the greater the rate decrease. When the oxide film removal process is violent and the fluctuation frequency is high, the pressing rate is rapidly reduced to provide sufficient time for oxide film removal. When the oxide film is gradually removed and the fluctuation frequency decreases, the rate decrease slows down, and the pressing process is smoothly advanced.

[0040] In this embodiment, the stage switching control method of the pressing process includes: continuously calculating the rate of change of pressing displacement per unit time during the initial frictional heat generation stage. The calculation of the rate of change of pressing displacement is achieved by the following formula:

[0041] in, The rate of change of the downward displacement. for The incremental downward displacement within. The calculation period for the displacement change rate is set between 10ms and 50ms to ensure the real-time performance and stability of the calculation results. When the downward displacement change rate is lower than a first change rate threshold and remains below it for a first set time, a stage switching command is generated. The first change rate threshold is a preset constant value, which corresponds to the baseline value of the displacement change rate under stable plastic flow conditions after the oxide film on the steel plate is completely removed. The first set time is set between 100ms and 500ms to avoid erroneous switching caused by instantaneous displacement fluctuations. According to the stage switching command, the downward pressure state of the rotating tool is switched from the initial frictional heating stage to the interface diffusion reaction stage, ensuring that the stage switching timing is perfectly matched with the actual state of the material. After the oxide film is completely removed and the plastic flow of the material is stable, the interface diffusion reaction stage is then entered.

[0042] Table 2. Displacement-pressure mapping relationship and torque correction parameters during the initial frictional heat generation stage. Compression depth range (mm) Displacement-pressure mapping coefficient kh Torque-pressure compensation coefficient kT Rate attenuation coefficient kv Initial displacement rate reference value (mm / s) 0-0.2 80 150 0.008 1.2 0.2-0.5 60 120 0.006 1.0 0.5-0.8 40 90 0.004 0.8 0.8-1.0 20 60 0.002 0.6 The table above clarifies the core control parameters corresponding to different compression depth ranges within the initial frictional heating stage, providing a quantitative basis for displacement-pressure mapping calculation, torque correction compensation, and nonlinear decay control of displacement rate. This ensures that the pressure and rate control in the initial frictional heating stage can be adapted to the oxide film removal state and plastic flow characteristics of materials at different compression depths, achieving uniform and complete oxide film removal.

[0043] In this embodiment, the increase in operating speed and the start of the initial frictional heating stage are triggered by the displacement offset detection during the pre-compression stage, ensuring the consistency of the initial state of the frictional heating process. The initial base pressure value is obtained by displacement pressure mapping based on real-time compression depth, and the pressure compensation value calculated by combining the root mean square value of the high-frequency torque fluctuation component is used to correct the initial base pressure value, generating a dynamically adjusted first dynamic pressure curve. At the same time, a nonlinear decreasing first displacement rate positively correlated with the torque fluctuation frequency is adopted, so that the mechanical energy input of the initial frictional heating stage can be adapted in real time to the process of removing the oxide film on the steel plate surface and the change of the material's plastic flow state, realizing the uniform removal of the oxide film on the steel plate surface and providing a stable contact interface for the subsequent interface diffusion reaction stage. The automatic switching of stages is achieved by continuous monitoring of the compression displacement change rate, ensuring the matching of stage switching timing with the material state.

[0044] refer to Figure 4 In a preferred embodiment, during the interface diffusion reaction stage, the real-time distribution data of the temperature field at the steel-aluminum interface is acquired by: extracting an infrared thermal radiation image sequence from the side of the lap joint at a preset sampling frequency. The infrared thermal radiation image sequence is acquired by a high-speed infrared thermal imager with a sampling frequency of not less than 200Hz and a spectral response range covering the mid-infrared band, which can accurately capture the thermal radiation signal of the steel-aluminum interface. The lens optical axis of the high-speed infrared thermal imager is set perpendicular to the side of the lap joint to ensure that the complete area of ​​the interface between the steel plate and the aluminum plate can be clearly acquired. Each frame of the acquired infrared thermal radiation image sequence has the same spatial resolution and time reference, ensuring the continuity and accuracy of subsequent temperature field calculations.

[0045] Gaussian filtering is performed on each frame of the infrared thermal radiation image sequence. The convolution kernel size of the Gaussian filter is set to 3×3 or 5×5, and the standard deviation is set between 1.0 and 1.5. This filters out Gaussian noise and salt-and-pepper noise generated during image acquisition, improving the signal-to-noise ratio of the image and avoiding noise interference with subsequent interface contour extraction and grayscale calculation. For each frame of the denoised infrared thermal radiation image, an edge detection operator is used to extract the interface contour line between the steel plate and the aluminum plate. The edge detection operator uses either the Sobel operator or the Canny operator. By calculating the gradient in the horizontal and vertical directions of the image, the grayscale abrupt change region between the steel plate and the aluminum plate is extracted. This grayscale abrupt change region is the interface contour line, which is a continuous line segment in the horizontal direction, corresponding to the physical bonding interface of dissimilar metals such as steel and aluminum. Based on the interface contour line, a rectangular region of interest is defined in each frame of the infrared thermal radiation image. The upper and lower boundaries of the rectangular region of interest are located at preset pixel distances above and below the interface contour line, respectively, and the left and right boundaries cover the entire welding influence area of ​​the lap joint. Subsequent grayscale processing and grayscale feature matrix extraction operations are performed only on the pixels within the rectangular region of interest, eliminating the interference of pixels in non-interface areas on the temperature field calculation and improving the calculation efficiency and accuracy of the temperature field data.

[0046] Pixels within the rectangular region of interest are converted to grayscale, transforming multi-channel infrared thermal radiation image data into single-channel grayscale values. The grayscale values ​​range from 0 to 255, corresponding to different thermal radiation intensities. The grayscale value is positively correlated with the temperature of the target area. The grayscale values ​​of all pixels within the rectangular region of interest are extracted and arranged according to their spatial positions to generate a grayscale feature matrix. The rows and columns of the grayscale feature matrix correspond to the vertical and horizontal spatial positions of pixels within the rectangular region of interest, respectively. Each element in the matrix represents the grayscale value of the pixel at its corresponding position. The grayscale feature matrix completely preserves the spatial distribution information of thermal radiation in the steel-aluminum interface region.

[0047] The grayscale feature matrix is ​​compared with a pre-calibrated temperature grayscale lookup table to generate real-time distribution data of the temperature field at the steel-aluminum interface. Specifically, the conversion from grayscale values ​​to temperature values ​​is achieved through the following mapping formula:

[0048] in, Coordinates within the steel-aluminum interface Real-time temperature value at the location coordinates within the grayscale feature matrix The grayscale value at the location, A pre-calibrated temperature-grayscale mapping function was used, obtained through calibration experiments. In these experiments, samples of the same material and thickness as the steel and aluminum plates used in welding were placed in a heating furnace and heated with a preset temperature gradient. Simultaneously, infrared thermal radiation images of the samples at different temperatures were acquired using a high-speed infrared thermal imager, identical to that used in the welding process. Grayscale values ​​at the corresponding temperatures were extracted to generate a temperature-grayscale lookup table. The temperature range in the table covers the highest temperature interval at the steel-aluminum interface during welding, from room temperature to 50°C above the critical temperature for plastic deformation of steel (not exceeding the melting point of steel). The real-time temperature field distribution data at the steel-aluminum interface is a two-dimensional matrix containing temperature values ​​at various spatial locations within the interface. The spatial resolution of the matrix is ​​consistent with the pixel resolution of the infrared thermal radiation image, fully reflecting the spatial temperature distribution at the steel-aluminum interface.

[0049] The output pressure value of the second dynamic pressure curve is piecewise linearly inversely adjusted according to the change of the highest temperature value in the real-time distribution data. The highest temperature value is the maximum value in the real-time distribution data of the temperature field at the steel-aluminum interface, corresponding to the hottest region at the interface. The temperature in this region directly determines the diffusion rate of steel and aluminum atoms and the growth rate of intermetallic compounds. The specific logic of the piecewise linearly inverse adjustment is implemented through the following formula:

[0050] in, This is the real-time output pressure value of the second dynamic pressure curve. This represents the highest temperature value in the real-time temperature field distribution data at the steel-aluminum interface. , , Three pre-set temperature thresholds, among which This is the critical temperature for plastic deformation of steel. This refers to the critical temperature for solid-state welding of steel. This is the preset maximum safe temperature for the interface. This is the initial pressure value at the start of the interfacial diffusion reaction stage, which is equal to the instantaneous pressure value at the end of the initial frictional heat generation stage. , This represents the pressure adjustment coefficient within the two temperature ranges. for The corresponding pressure value at that time This is the preset minimum pressure value to avoid insufficient frictional heat input due to excessively low pressure.

[0051] In this piecewise linear inverse proportional adjustment logic, when the maximum interface temperature increases, the output pressure value is reduced to decrease the frictional heat input and avoid excessive growth of intermetallic compounds due to excessively high interface temperature; when the maximum interface temperature decreases, the output pressure value is increased to increase the frictional heat input, ensuring that the steel material at the interface maintains a stable plastic flow state, providing sufficient energy and contact conditions for the diffusion reaction of steel and aluminum atoms, thus realizing closed-loop control of interface heat input.

[0052] In this embodiment, the stage switching control method of the pressing process further includes: during the interface diffusion reaction stage, when the area ratio of the critical temperature region for plastic softening on the steel side in the real-time distribution data reaches a preset area ratio threshold, a second stage switching command is generated and the process switches to the upsetting stage. The calculation method for the area ratio of the critical temperature region for plastic softening on the steel side is as follows: in the real-time distribution data of the temperature field at the steel-aluminum interface, the area of ​​the region with a temperature value higher than the critical temperature for plastic deformation of steel is counted, and this area is divided by the total area of ​​the rectangular region of interest to obtain the area ratio of the plastic softening region on the steel side. The preset area ratio threshold is set between 30% and 60% to ensure that a sufficient plastic softening region is formed on the steel side, providing sufficient conditions for the interface diffusion reaction, while avoiding excessive material splashing during the subsequent upsetting process due to an excessively large plastic softening region. When the area ratio of the plastic softening region on the steel side reaches the preset threshold, it indicates that the interface diffusion reaction has reached the expected state. At this time, a second stage switching command is generated to switch the pressing state of the rotating tool from the interface diffusion reaction stage to the upsetting stage, ensuring that the stage switching timing is completely matched with the interface metallurgical reaction state.

[0053] Table 3. Comparison of Gray Scale Calibration for Steel-Aluminum Interface Temperature Temperature value (°C) Corresponding to 8-bit grayscale Allowable fluctuation range of grayscale values 25 20 18-22 100 45 42-48 200 78 75-81 300 112 109-115 400 146 143-149 500 180 177-183 600 214 211-217 700 235 232-238 800 248 245-250 The table above serves as the core calibration basis for calculating the temperature field at the steel-aluminum interface. It clarifies the range of grayscale values ​​in infrared thermal radiation images corresponding to different interface temperatures. This table allows for the rapid conversion of the grayscale feature matrix into real-time distribution data of the temperature field at the steel-aluminum interface, ensuring the accuracy and real-time nature of the temperature field data and providing a reliable input basis for adjusting the second dynamic pressure curve.

[0054] In this embodiment, Gaussian filtering denoising improves the signal-to-noise ratio of the infrared thermal radiation image. An edge detection operator is used to extract the outline of the steel-aluminum interface and delineate a rectangular region of interest. Subsequent grayscale processing and temperature calculation are performed only on this region, eliminating interference from non-interface areas and improving the computational efficiency and accuracy of the temperature field data. A pre-calibrated temperature grayscale reference table is used to convert the grayscale feature matrix into real-time distribution data of the steel-aluminum interface temperature field, realizing non-contact real-time detection of the interface temperature field. A second dynamic pressure curve, which is piecewise linearly inversely proportional to the highest interface temperature value, achieves closed-loop coupling control of the pressure and interface temperature field during the interface diffusion reaction stage. This allows the heat input of the interface to be adjusted in real time with changes in interface temperature, effectively controlling the growth rate and thickness of intermetallic compounds at the interface and providing stable temperature and pressure conditions for the diffusion reaction of steel and aluminum atoms. Automatic switching of stages is achieved by monitoring the area ratio of the plastic softening region on the steel side, ensuring the matching of the end time of the interface diffusion reaction stage with the metallurgical reaction state of the interface.

[0055] refer to Figure 5 In a preferred embodiment, during the upsetting stage, the stepped pressure curve comprises three consecutive pressure steps, executed sequentially along the downsetting time. Each pressure step corresponds to a fixed pressure setpoint, and the pressure setpoints increase sequentially. The pressure value of the first pressure step is set to a first multiple of the instantaneous pressure value at the end of the interfacial diffusion reaction stage; the pressure value of the second pressure step is set to a second multiple of the instantaneous pressure value; and the pressure value of the third pressure step is set to a third multiple of the instantaneous pressure value. The first multiple, the second multiple, and the third multiple increase sequentially. The first multiple is set between 1.1 and 1.3, the second multiple is set between 1.4 and 1.6, and the third multiple is set between 1.7 and 2.0. The sequentially increasing pressure values ​​of the three pressure steps provide a gradually increasing upsetting force for the lap joint, promoting the densification of the plastic softening material at the interface and the formation of the weld nugget. The third displacement rate remains constant throughout the three consecutive pressure steps. The value of the third displacement rate is lower than that of the second displacement rate at the end of the interfacial diffusion reaction stage, ensuring that the downward displacement of the rotating tool during upsetting is stable and controllable, and avoiding weld nugget formation defects and material spatter caused by displacement rate fluctuations.

[0056] Within each of the three consecutive pressure steps, the normal pressure value applied to the lap joint by the rotating tool and the real-time indentation depth value of the rotating tool are collected in real time. The normal pressure value is collected in real time by a pressure sensor built into the spindle, and the real-time indentation depth value is collected in real time by a displacement sensor built into the spindle. The collection frequency is no less than 1kHz to ensure the real-time performance and accuracy of the data. The ratio of the normal pressure value to the real-time indentation depth value is calculated to obtain the real-time deformation resistance value. The real-time deformation resistance value directly reflects the plastic deformation resistance of the lap joint material under the current compression state. Its value change corresponds to the densification process and hardening state of the material, specifically achieved through the following formula:

[0057] in, The real-time deformation resistance value at time t. The normal pressure value collected at time t. The real-time indentation depth value at time t is the axial displacement value of the end face of the rotating tool relative to the initial upper surface of the steel plate. It uses the same reference coordinate system as the indentation depth calculation in the previous stage to ensure the continuity of the values.

[0058] When the increase in the real-time deformation resistance value compared to the deformation resistance value of the previous acquisition cycle exceeds a preset resistance increase threshold, a step jump command is triggered, and the system proceeds to the next pressure step according to the step jump command. The increase in the deformation resistance value is calculated using the following formula:

[0059] in, This represents the increase in real-time deformation resistance value. This represents the deformation resistance value from the previous data collection cycle. The preset resistance increase threshold is set between 5% and 15%. When the increase in deformation resistance exceeds this threshold, it indicates that the material's plastic deformation resistance has increased, and the current pressure value is insufficient to meet the material densification requirements. An increase in the upsetting force is needed to continue achieving material densification and weld nugget formation. At this point, a step jump is triggered, entering the next higher pressure step. Within each pressure step, if a step jump command is not triggered, the pressure value and third displacement rate of the current pressure step are maintained, continuing downward pressure until a step jump command is triggered or the preset maximum displacement value corresponding to that pressure step is reached. The preset maximum displacement value is used to avoid excessive downward displacement within a single pressure step, which could lead to weld nugget formation defects.

[0060] refer to Figure 6The pressure holding operation during the upsetting stage includes: after the rotating tool presses down to the final welding depth according to the stepped pressure curve, a pressure holding operation is performed. The final welding depth is the preset maximum pressing depth, which does not exceed 60% of the steel plate thickness, to avoid direct contact between the rotating tool and the aluminum plate, thus preventing tool wear and interface defects. The pressure holding operation is performed during the maintenance of the third pressure step. During the pressure holding process, the pressing depth of the rotating tool remains unchanged at the final welding depth, the rotation speed of the rotating tool remains unchanged at the set operating speed, and the pressure value remains unchanged at the pressure value of the third pressure step. By maintaining constant pressure and depth, the material inside the weld nugget is densified, eliminating forming defects such as porosity and shrinkage cavities.

[0061] During the maintenance of the third pressure step, the variance of all real-time deformation resistance values ​​within the sliding window is continuously calculated using a second set time as the sliding window. The variance directly reflects the degree of fluctuation of the deformation resistance value, corresponding to the stable state of the material's plastic deformation, and is specifically achieved through the following formula:

[0062] in, M represents the variance of the real-time deformation resistance values ​​within the sliding window, and M is the number of sampling points for the deformation resistance values ​​within the sliding window. Let j be the deformation resistance value of the j-th sampling point within the sliding window. This is the average value of all deformation resistance values ​​within the sliding window. The second set time is set between 200ms and 1000ms, and the sliding window slides continuously with the sampling time to ensure that fluctuations in deformation resistance values ​​can be monitored in real time.

[0063] The variance is compared with a preset variance threshold. When the variance is less than the preset variance threshold for a third consecutive set time, a pressure holding termination command is generated. The third set time is set between 100ms and 500ms. The preset variance threshold is a pre-set constant value, corresponding to the deformation resistance fluctuation level after the material's plastic deformation stabilizes and the weld nugget densifies. When the variance is less than the preset variance threshold for a third consecutive set time, it indicates that the weld nugget densification process is complete and the material's plastic deformation has entered a stable state. At this time, a pressure holding termination command is generated. Based on the pressure holding termination command, the rotating tool is controlled to unload the pressure and lift upward to detach from the lap joint, completing the entire needle-free friction spot welding process.

[0064] Table 4. Step pressure control parameters and jump conditions during the upsetting stage. Pressure step sequence number Pressure ratio Third displacement rate constant value (mm / s) Resistance increase threshold (%) Maximum displacement of a single step (mm) First step 1.2 0.3 10 0.15 Second Tier 1.5 0.3 8 0.10 Third step 1.9 0.3 5 0.05 The table above clarifies the core control parameters and jump trigger conditions of the three continuous pressure steps in the upsetting forming stage, providing a quantitative basis for the execution of the step-increasing pressure curve and the automatic jump of the pressure steps, ensuring that the pressure applied in the upsetting process can be adapted to the actual deformation state of the material in real time, and realizing the stable and dense forming of the weld nugget.

[0065] In this embodiment, a stepped pressure curve consisting of three continuously increasing pressure steps provides a gradually increasing upsetting force for the upsetting process. Combined with a constant third displacement rate throughout the process, this ensures the smooth and controllable plastic deformation of the material during upsetting. The real-time deformation resistance value is obtained by calculating the ratio of the normal pressure value to the real-time indentation depth value. The pressure step jump is triggered based on the increase in the deformation resistance value, making the increase in upsetting force directly related to the actual deformation resistance of the material, thus realizing adaptive control of the upsetting process. The variance of the deformation resistance value is continuously calculated through a sliding window. The termination of the pressure holding operation is controlled based on the continuous stable state of the variance, ensuring that the pressure holding process continues until the weld nugget is completely densified and the plastic deformation of the material is completely stable. This effectively improves the forming quality of the weld nugget and the mechanical properties of the joint.

Claims

1. A needle-free friction spot welding method for dissimilar metals of aluminum and steel based on multi-stage pressure, characterized in that, include: A steel plate is placed on top of an aluminum plate to form a lap joint, and a rotating tool is started to frictionally heat the lap joint. The pressing process of the rotary tool is divided into an initial frictional heat generation stage, an interfacial diffusion reaction stage, and an upsetting stage. During the initial frictional heat generation stage, the rotating tool is pressed down according to the first displacement rate and the first dynamic pressure curve until the first preset displacement threshold is reached. During the interface diffusion reaction stage, the rotating tool is pressed down according to the second displacement rate and the second dynamic pressure curve until the second preset displacement threshold is reached. The output pressure value of the second dynamic pressure curve is coupled with the real-time distribution data of the temperature field at the steel-aluminum interface. During the upsetting stage, the rotary tool is pressed down to the final welding depth according to the third displacement rate and the step-increasing pressure curve, and a pressure holding operation is performed at the final welding depth.

2. The needleless friction spot welding method for dissimilar metals of aluminum and steel based on multi-stage downward pressure according to claim 1, characterized in that, In the initial frictional heat generation stage, the generation process of the first dynamic pressure curve includes: acquiring the real-time compression depth value of the rotating tool, inputting the real-time compression depth value into a preset displacement-pressure mapping relationship model, and calculating the initial base pressure value; collecting the real-time torque value of the rotating tool during rotation, correcting the initial base pressure value based on the real-time torque value, and generating the first dynamic pressure curve; the first displacement rate is set as a nonlinear deceleration rate, and the decrease amplitude of the nonlinear deceleration rate is positively correlated with the fluctuation frequency of the real-time torque value.

3. The needleless friction spot welding method for dissimilar metals of aluminum and steel based on multi-stage downward pressure according to claim 1, characterized in that, In the interface diffusion reaction stage, the real-time distribution data of the steel-aluminum interface temperature field is obtained by: extracting an infrared thermal radiation image sequence from the side of the lap joint at a preset sampling frequency; performing grayscale processing on the infrared thermal radiation image sequence to extract the grayscale feature matrix of the steel-aluminum interface; comparing the grayscale feature matrix with a pre-calibrated temperature grayscale reference table to generate real-time distribution data of the steel-aluminum interface temperature field; and adjusting the output pressure value of the second dynamic pressure curve in a piecewise linear inverse proportion to the change of the highest temperature value in the real-time distribution data.

4. The needleless friction spot welding method for dissimilar metals of aluminum and steel based on multi-stage downward pressure according to claim 1, characterized in that, In the upsetting stage, the stepped pressure curve includes three consecutive pressure steps. The pressure value of the first pressure step is set to a first multiple of the instantaneous pressure value at the end of the interfacial diffusion reaction stage. The pressure value of the second pressure step is set to a second multiple of the instantaneous pressure value. The pressure value of the third pressure step is set to a third multiple of the instantaneous pressure value. The first multiple, the second multiple, and the third multiple increase sequentially. The third displacement rate remains constant within the three consecutive pressure steps.

5. The needleless friction spot welding method for dissimilar metals of aluminum and steel based on multi-stage pressure as described in claim 1, characterized in that, Before starting the rotating tool to perform friction heating on the lap joint, the method further includes: applying a pre-clamping force to the rotating tool so that the end face of the rotating tool contacts the surface of the steel plate; during the application of the pre-clamping force, controlling the rotating tool to rotate at an initial speed, and simultaneously recording the axial displacement offset of the spindle of the rotating tool; when the axial displacement offset of the spindle reaches a preset offset threshold, increasing the speed of the rotating tool to a set working speed, and simultaneously triggering the initial friction heating stage.

6. The needleless friction spot welding method for dissimilar metals of aluminum and steel based on multi-stage downward pressure according to claim 1, characterized in that, The stage switching control method of the pressing process includes: during the initial frictional heating stage, continuously calculating the rate of change of pressing displacement per unit time; when the rate of change of pressing displacement is lower than a first rate of change threshold and continues for a first set time, generating a stage switching command; according to the stage switching command, switching the pressing state of the rotating tool from the initial frictional heating stage to the interface diffusion reaction stage; during the interface diffusion reaction stage, when the area ratio of the critical temperature region for plastic softening on the steel side in the real-time distribution data reaches a preset area ratio threshold, generating a second stage switching command and switching to the upsetting forming stage.

7. The needleless friction spot welding method for dissimilar metals of aluminum and steel based on multi-stage pressure as described in claim 2, characterized in that, The process of correcting the initial base pressure value based on the real-time torque value includes: extracting the high-frequency torque fluctuation component from the real-time torque value and calculating the root mean square value of the high-frequency torque fluctuation component; dividing the root mean square value by the average value of the real-time torque value to obtain the torque fluctuation coefficient; inputting the torque fluctuation coefficient into a preset pressure compensation function to calculate the pressure compensation amount; adding the pressure compensation amount to the initial base pressure value to generate the corrected pressure value, and the corrected pressure value constitutes the first dynamic pressure curve.

8. The needleless friction spot welding method for dissimilar metals of aluminum and steel based on multi-stage pressure as described in claim 3, characterized in that, Before performing grayscale processing on the infrared thermal radiation image sequence, the method further includes: performing Gaussian filtering denoising processing on each frame of the infrared thermal radiation image sequence; for each frame of the denoised infrared thermal radiation image, using an edge detection operator to extract the contour line of the interface between the steel plate and the aluminum plate; defining a rectangular region of interest in each frame of the infrared thermal radiation image based on the interface contour line, and performing subsequent grayscale processing and grayscale feature matrix extraction operations only on the pixels within the rectangular region of interest.

9. The needleless friction spot welding method for aluminum and steel dissimilar metals based on multi-stage downward pressure according to claim 4, characterized in that, In each of the three consecutive pressure steps, the normal pressure value applied by the rotating tool to the lap joint and the real-time indentation depth value of the rotating tool are collected in real time; the ratio of the normal pressure value to the real-time indentation depth value is calculated to obtain the real-time deformation resistance value. When the increase in the real-time deformation resistance value compared to the deformation resistance value of the previous acquisition cycle exceeds the preset resistance increase threshold, a step jump command is triggered, and the next pressure step is entered according to the step jump command.

10. The needleless friction spot welding method for aluminum and steel dissimilar metals based on multi-stage pressure as described in claim 9, characterized in that, The process of performing the pressure holding operation during the upsetting stage includes: during the maintenance of the third pressure step, using a second set time as a sliding window, continuously calculating the variance of all the real-time deformation resistance values ​​within the sliding window; comparing the variance with a preset variance threshold, and generating a pressure holding termination command when the variance is less than the preset variance threshold for a third consecutive set time; and controlling the rotating tool to unload the pressure and lift it upwards to disengage from the lap joint according to the pressure holding termination command.

Images